Sputtering processes are typically performed in a vacuum chamber. The chamber includes a target comprised of material, such as Aluminum, to be sputtered. A substrate, such as a semiconductor wafer bearing integrated circuits, is placed in the chamber and the chamber is evacuated. Once the chamber is evacuated, a process gas is introduced into the chamber at a low pressure, and a voltage is applied to the target. Ionized gas in the chamber is accelerated by the electric fields of the target. When the ions impinge upon the target, atoms of sputtering material are dislodged ("sputtered") from the target. The dislodged atoms deposit on the substrate, forming, over time, a thin film of target material on the substrate.
The sputtering process described above slowly wears material from the target until, ultimately, the target must be renewed. Typically, this is done by removing the target from the chamber and inserting a new target into the chamber.
FIG. 1 is a cross-sectional view of a typical vacuum chamber 10 for performing a sputtering process, and a target assembly 12 mounted in this chamber. Chamber 10 is generally cylindrical or rectangular in shape, and target 12 is generally disk-shaped and mounts into a circular opening 13 in chamber 10. Wafer 14, which is disk-shaped, is supported inside of chamber 10 by a generally disk-shaped anode 16. Electrical power is applied to the anode 16 with respect to the remainder of the chamber 10.
As shown in FIG. 1, a typical target is manufactured of a backing plate 20 of a metal such as Copper, metallurgically bonded (e.g., soldered) to a target plate having a front face 22 of sputtering material. Backing plate 20 is affixed to a chamber cover 23 by mounting screws 24. Cover 23 is clamped to chamber 10 by means not shown, and supports the front face 22 in a position generally opposite wafer 14. An O-ring inserted into groove 26 prevents leakage into the chamber 10 through the target-chamber interface. During the sputtering process, material from front face 22 is removed and deposited on the wafer 14, while backing plate remains in place.
Over time, a typical target such as that shown in FIG. 1 wears to the extent that the sputtering material on the front face 22 must be renewed. When this occurs, cover 23 is lifted from chamber 10, and target assembly 12 is detached from cover 23 and removed. A new target assembly 12 is then assembled to cover 23, and cover 23 is clamped onto chamber 10, and processing continues.
Backing plates are typically manufactured of expensive metals, and for this reason are typically recycled after use. Once a worn target assembly has been removed from the cover 23, the backing plate and the remaining sputtering material soldered thereto, are returned to the manufacturer. The manufacturer removes the remaining sputtering material from the backing plate, and solders a new target plate having a front face 22 of sputtering material to the backing plate. The target assembly 12, consisting of the recycled backing plate and new front face of sputtering material, is then ready for resale. To ensure that processing facilities promptly return used target assemblies to the manufacturer, manufacturers typically include a core charge in the price of each target sold; this charge is refunded upon return of the used target to the manufacturer.
This process for recycling used targets is inconvenient in that it requires regular shipment of bulky and heavy targets back to the manufacturer. Furthermore, to ensure a sufficient stockpile of targets to avoid downtime, semiconductor fabrication facilities must retain in inventory a number of new target assemblies 12, which involves essentially permanent payment of the core charges associated with the target assemblies held in inventory.